An elastic model is developed to investigate the initiation of normal and thrust faults in thrust wedges. The model assumes frictional sliding along the base and a linear variation of shear and normal tractions with depth along the rear of the thrust wedge. Using this model, the roles of basal friction, pore fluid pressure, and wedge configuration in controlling the mechanics of thrust wedges were evaluated. The model predicts coeval development of normal and thrust faults in the same thrust wedge and a listric shape for both thrust and normal faults. In particular, lower friction favors dominantly horizontal compression in a thrust wedge, whereas higher friction can produce dominantly horizontal extension. Wedges bounded by steep thrusts (≥45¿) are more likely to experience extension regardless of their sizes and the boundary conditions applied. For wedges bounded by shallow-dipping thrusts (≤15¿) and under the same boundary conditions, shorter wedges are more dominated by compression than longer wedges. Using this model, the length of a Hubbert-Rubey thrust toe (defined as the length of the unfractured, frontal portion of a thrust wedge) is calculated. With the same boundary conditions and mechanical properties, a thinner wedge, in general, favors a longer Hubbert-Rubey toe than a thicker wedge. This simple model is applied to explain the initiation of the short-lived Miocene normal fault system in the Higher Himalaya as a consequence of a rapid release of pore fluid pressure along the Main Central Thrust during the formation of two-mica leucogranites in the hanging wall. It also stimulates the occurrence of normal fault associated with the El Asnam thrust-type earthquake. ¿ American Geophysical Union 1993